Run a Scientific Review of a Task
Tasks
Run a scientific review
Have a reviewer agent double-check a completed Task’s analysis for scientific accuracy.
POST
Run a Scientific Review of a Task
A reviewer agent re-reads a completed Task’s analysis and checks it for
scientific accuracy, correct use of the data, unsupported claims, and stated
limitations. Reviewing an
idle or still-running Task has nothing to look at
yet.
It runs in the background. This call returns immediately with
review_status: "running" and does the work off the request path — it does
not return the review itself.
Get the result by polling GET /tasks/{task_id}: watch review_status go
running → done | failed. When it’s done, the review is appended to the Task’s
conversation as a new message — read it with GET /messages?task_id=.... On
failed, review_error explains why.
Re-running: calling this while a review is already running is a no-op (you
get the in-flight running state back). After a review finishes, calling again
starts a fresh one that overwrites the previous result.